Chemistry beyond the scale of exact diagonalization on a quantum-centric supercomputer

Area of Science:

• Quantum computing
• Computational chemistry
• Supercomputing

Background:

• Universal quantum computers can simulate quantum systems, but face runtime challenges for practical applications like electronic structure simulations.
• Current quantum processors are approaching the necessary scale (hundreds of qubits), yet deep circuits and extensive measurements limit their standalone utility.

Purpose of the Study:

• To demonstrate a hybrid quantum-classical workflow for electronic structure simulations, offloading computational tasks to classical distributed computing.
• To address the prohibitive runtimes of quantum computers for complex chemistry problems.

Main Methods:

• Utilized a Heron superconducting processor and the Fugaku supercomputer for a distributed computing approach.
• Developed an algorithm to process quantum samples, generating ground-state energy upper bounds and sparse wave function approximations.
• Simulated the ground-state dissociation of N₂ and properties of [2Fe-2S] and [4Fe-4S] clusters using circuits up to 77 qubits and 10,570 gates.

Main Results:

• Successfully simulated challenging electronic structure problems for N₂ and iron-sulfur clusters.
• Demonstrated that a quantum-centric supercomputing architecture can handle problems exceeding exact diagonalization capabilities.
• Generated upper bounds for ground-state energy and sparse approximations of wave functions.

Conclusions:

• A hybrid quantum-classical approach effectively overcomes the limitations of current quantum hardware for complex chemistry simulations.
• Quantum-centric supercomputing architectures show promise for tackling advanced computational chemistry problems.
• This method enables simulations beyond the scope of traditional exact diagonalization techniques for current quantum devices.